|
|
FAQ |
|
HLsyn Frequently Asked Questions (FAQ)Q. During installation, the HLsyn installer asks for a serial key. What should I enter?A. Please use "sensimetrics" (without quotes) as the serial key for HLsyn. Q. I'm getting this error (or something similar):"Error: a numerical or character function was given and argument xxxx/2 of wrong type HLSYN-MAIN-WINDOW-BOX"What should I do?A. A patch is available from or support page which solves problems due to non-standard and newer standard screen sizes. Please download the patch here. Q. Will HLsyn run on Windows® 95/98/Me/NT/2000/XP/Vista?A. Yes. Q. Why am I getting errors while trying to install/run HLsyn on Windows® 2000/XP/Vista? A. On operating systems based on NT 5 (Windows® 2000, Windows® XP, Windows® Vista) it is necessary to log in as an administrator to properly install HLsyn. While installing as a Power User may appear to succeed, in fact the program will not be correctly installed. Note: On Windows® XP all user accounts other than Limited Users normally have administrator privileges and may be used to install HLsyn. Q. Is HLsyn available for operating systems other than Windows?A. At this time, no. Presently, we only offer a Windows-based package. Q. Is HLsyn a text-to-speech system?A. No. HLsyn is a quasi-articulatory synthesizer; the user controls the speech output by adjusting a set of time-varying parameters and a set of constants describing a vocal-tract model. Q. What's the difference between HLsyn and SenSyn?A. HLsyn is a "high-level" parametric synthesizer that produces more natural-sounding speech from fewer time-varying parameters (13) than SenSyn (48). HLsyn uses fewer parameters by exploiting the interdependence of the parameters and subjecting them to inter-parameter constraints. These 13 parameters are more easily interpreted in terms of articulation. HLsyn transforms the HL parameters into the lower (Klatt) level (or KL) parameters of SenSyn through complex mapping relations. Additionally, HLsyn's user interface is much easier to use, and more user-friendly, than SenSyn's. Q. What are the HL parameters?A. There are thirteen (13) time-varying high-level parameters: five constriction area parameters, four formant frequency parameters, a fundamental frequency parameter, a subglottal pressure parameter, and two parameters dealing with motion and stiffness of the vocal tract walls.. There are also numerous HL speaker constants that are user-adjustable but not time-variable. For a chart of the HL parameters, click here. Q. Can I access the lower-level KL parameters from HLsyn?A. Yes. In addition to the thirteen high-level parameters of HLsyn, users may access and control the KL parameters. Q. Does HLsyn permit users to playback the synthesized speech or view a spectrogram of it?A. Yes. Users can playback the results of the synthesis (assuming they have an audio card), or analyze the synthesis files in the form of a spectrogram. Q. How can I import or export parameters from HLsyn?A. You can save your work as either of two types of file (.HLD, the usual format, and .KLD), and later read these files again in HLsyn. Also, you can export and import parameter tracks (all 13 HL parameters, all KL parameters, subsidiary pressure and flow parameters, and the speaker constants) as text files; this facility permits exchange of parameter data with other programs (with a little programming effort). Q. What is a .KLD file?A. These "Klatt Level Documents" are binary files which can be written and read by HLsyn. They contain the result of converting the high-level HL parameters to Klatt level parameters (hence the name) but contain no HL parameter information. Q. If I want to edit the lower-level (KL) parameters, do I have to edit all the HL parameters first?A. You can generate KL parameters by hand by creating a .KLD file. Alternatively, you can generate KL parameters automatically from HL parameters. You then need to save the KL parameter table in .KLD file form after which you can edit the underlying lower-level parameters. (By saving as a .KLD file you have removed the HL parameter information. If you do not save as .KLD then the next time you synthesize your utterance HLsyn will use the still-present HL parameters to regenerate all the KL parameters, thus obliterating your edits.) Q. Can you recommend reading materials on the mapping relation between high-level and low-level parameters? Klatt synthesizers? Etc.?A. Yes. Q. Do I need any additional hardware to run HLsyn?A. No. All you need is a Windows-compatible audio card to play back your synthesis.
|
|